[{"publication_identifier":{"issn":["1057-7149"],"eissn":["1941-0042"]},"page":"5707-5722","arxiv":1,"day":"01","fulldoi":"https://doi.org/10.1109/tip.2018.2859044","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","_id":"18271","title":"Class-aware fully convolutional Gaussian and Poisson denoising","external_id":{"arxiv":["1808.06562"]},"article_type":"original","article_processing_charge":"No","volume":27,"main_file_link":[{"url":"https://doi.org/10.48550/arXiv.1808.06562","open_access":"1"}],"doi":"10.1109/tip.2018.2859044","oa":1,"issue":"11","author":[{"first_name":"Tal","last_name":"Remez","full_name":"Remez, Tal"},{"full_name":"Litany, Or","first_name":"Or","last_name":"Litany"},{"first_name":"Raja","last_name":"Giryes","full_name":"Giryes, Raja"},{"last_name":"Bronstein","first_name":"Alexander","full_name":"Bronstein, Alexander","id":"58f3726e-7cba-11ef-ad8b-e6e8cb3904e6","orcid":"0000-0001-9699-8730"}],"scopus_import":"1","OA_type":"green","date_updated":"2024-10-16T13:00:30Z","status":"public","quality_controlled":"1","OA_place":"repository","type":"journal_article","language":[{"iso":"eng"}],"abstract":[{"lang":"eng","text":"We propose a fully convolutional neural-network architecture for image denoising which is simple yet powerful. Its structure allows to exploit the gradual nature of the denoising process, in which the shallow layers handle local noise statistics, while deeper layers recover edges and enhance textures. Our method advances the state of the art when trained for different noise levels and distributions (both Gaussian and Poisson). In addition, we show that making the denoiser class-aware by exploiting semantic class information boosts the performance, enhances the textures, and reduces the artifacts."}],"publisher":"Institute of Electrical and Electronics Engineers","extern":"1","month":"11","date_published":"2018-11-01T00:00:00Z","oa_version":"Preprint","citation":{"mla":"Remez, Tal, et al. “Class-Aware Fully Convolutional Gaussian and Poisson Denoising.” <i>IEEE Transactions on Image Processing</i>, vol. 27, no. 11, Institute of Electrical and Electronics Engineers, 2018, pp. 5707–22, doi:<a href=\"https://doi.org/10.1109/tip.2018.2859044\">10.1109/tip.2018.2859044</a>.","chicago":"Remez, Tal, Or Litany, Raja Giryes, and Alex M. Bronstein. “Class-Aware Fully Convolutional Gaussian and Poisson Denoising.” <i>IEEE Transactions on Image Processing</i>. Institute of Electrical and Electronics Engineers, 2018. <a href=\"https://doi.org/10.1109/tip.2018.2859044\">https://doi.org/10.1109/tip.2018.2859044</a>.","apa":"Remez, T., Litany, O., Giryes, R., &#38; Bronstein, A. M. (2018). Class-aware fully convolutional Gaussian and Poisson denoising. <i>IEEE Transactions on Image Processing</i>. Institute of Electrical and Electronics Engineers. <a href=\"https://doi.org/10.1109/tip.2018.2859044\">https://doi.org/10.1109/tip.2018.2859044</a>","short":"T. Remez, O. Litany, R. Giryes, A.M. Bronstein, IEEE Transactions on Image Processing 27 (2018) 5707–5722.","ista":"Remez T, Litany O, Giryes R, Bronstein AM. 2018. Class-aware fully convolutional Gaussian and Poisson denoising. IEEE Transactions on Image Processing. 27(11), 5707–5722.","ieee":"T. Remez, O. Litany, R. Giryes, and A. M. Bronstein, “Class-aware fully convolutional Gaussian and Poisson denoising,” <i>IEEE Transactions on Image Processing</i>, vol. 27, no. 11. Institute of Electrical and Electronics Engineers, pp. 5707–5722, 2018.","ama":"Remez T, Litany O, Giryes R, Bronstein AM. Class-aware fully convolutional Gaussian and Poisson denoising. <i>IEEE Transactions on Image Processing</i>. 2018;27(11):5707-5722. doi:<a href=\"https://doi.org/10.1109/tip.2018.2859044\">10.1109/tip.2018.2859044</a>"},"publication_status":"published","year":"2018","date_created":"2024-10-09T07:42:49Z","publication":"IEEE Transactions on Image Processing","intvolume":"        27"},{"volume":15,"main_file_link":[{"url":"http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:hbz:386-kluedo-14003","open_access":"1"}],"article_processing_charge":"No","article_type":"original","external_id":{"pmid":["17076408 "]},"issue":"11","doi":" 10.1109/TIP.2006.877501 ","oa":1,"fulldoi":"https://doi.org/ 10.1109/TIP.2006.877501 ","page":"3501 - 3513","publication_identifier":{"issn":["1057-7149"],"eissn":["1941-0042"]},"day":"01","title":"An optimal non-orthogonal separation of the anisotropic Gaussian convolution filter","_id":"3695","user_id":"317138e5-6ab7-11ef-aa6d-ffef3953e345","oa_version":"None","citation":{"mla":"Lampert, Christoph, and Oliver Wirjadi. “An Optimal Non-Orthogonal Separation of the Anisotropic Gaussian Convolution Filter.” <i>IEEE Transactions on Image Processing (TIP)</i>, vol. 15, no. 11, IEEE, 2006, pp. 3501–13, doi:<a href=\"https://doi.org/ 10.1109/TIP.2006.877501 \"> 10.1109/TIP.2006.877501 </a>.","chicago":"Lampert, Christoph, and Oliver Wirjadi. “An Optimal Non-Orthogonal Separation of the Anisotropic Gaussian Convolution Filter.” <i>IEEE Transactions on Image Processing (TIP)</i>. IEEE, 2006. <a href=\"https://doi.org/ 10.1109/TIP.2006.877501 \">https://doi.org/ 10.1109/TIP.2006.877501 </a>.","apa":"Lampert, C., &#38; Wirjadi, O. (2006). An optimal non-orthogonal separation of the anisotropic Gaussian convolution filter. <i>IEEE Transactions on Image Processing (TIP)</i>. IEEE. <a href=\"https://doi.org/ 10.1109/TIP.2006.877501 \">https://doi.org/ 10.1109/TIP.2006.877501 </a>","short":"C. Lampert, O. Wirjadi, IEEE Transactions on Image Processing (TIP) 15 (2006) 3501–3513.","ista":"Lampert C, Wirjadi O. 2006. An optimal non-orthogonal separation of the anisotropic Gaussian convolution filter. IEEE Transactions on Image Processing (TIP). 15(11), 3501–3513.","ieee":"C. Lampert and O. Wirjadi, “An optimal non-orthogonal separation of the anisotropic Gaussian convolution filter,” <i>IEEE Transactions on Image Processing (TIP)</i>, vol. 15, no. 11. IEEE, pp. 3501–3513, 2006.","ama":"Lampert C, Wirjadi O. An optimal non-orthogonal separation of the anisotropic Gaussian convolution filter. <i>IEEE Transactions on Image Processing (TIP)</i>. 2006;15(11):3501-3513. doi:<a href=\"https://doi.org/ 10.1109/TIP.2006.877501 \"> 10.1109/TIP.2006.877501 </a>"},"date_published":"2006-11-01T00:00:00Z","pmid":1,"publisher":"IEEE","month":"11","extern":"1","intvolume":"        15","publication_status":"published","publication":"IEEE Transactions on Image Processing (TIP)","date_created":"2018-12-11T12:04:40Z","year":"2006","OA_type":"closed access","publist_id":"2669","date_updated":"2026-08-27T11:04:53Z","author":[{"orcid":"0000-0001-8622-7887","first_name":"Christoph","last_name":"Lampert","id":"40C20FD2-F248-11E8-B48F-1D18A9856A87","full_name":"Lampert, Christoph"},{"full_name":"Wirjadi, Oliver","first_name":"Oliver","last_name":"Wirjadi"}],"language":[{"iso":"eng"}],"abstract":[{"lang":"eng","text":"We give an analytical and geometrical treatment of what it means to separate a Gaussian kernel along arbitrary axes in Ropfn, and we present a separation scheme that allows us to efficiently implement anisotropic Gaussian convolution filters for data of arbitrary dimensionality. Based on our previous analysis we show that this scheme is optimal with regard to the number of memory accesses and interpolation operations needed. The proposed method relies on nonorthogonal convolution axes and works completely in image space. Thus, it avoids the need for a fast Fourier transform (FFT)-subroutine. Depending on the accuracy and speed requirements, different interpolation schemes and methods to implement the one-dimensional Gaussian (finite impulse response and infinite impulse response) can be integrated. Special emphasis is put on analyzing the performance and accuracy of the new method. In particular, we show that without any special optimization of the source code, it can perform anisotropic Gaussian filtering faster than methods relying on the FFT."}],"type":"journal_article","status":"public"}]